From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 11099 invoked by alias); 22 Feb 2004 21:10:19 -0000 Mailing-List: contact gdb-patches-help@sources.redhat.com; run by ezmlm Precedence: bulk List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sources.redhat.com Received: (qmail 11070 invoked from network); 22 Feb 2004 21:10:18 -0000 Received: from unknown (HELO walton.kettenis.dyndns.org) (213.93.115.144) by sources.redhat.com with SMTP; 22 Feb 2004 21:10:18 -0000 Received: from elgar.kettenis.dyndns.org (elgar.kettenis.dyndns.org [192.168.0.2]) by walton.kettenis.dyndns.org (8.12.6p3/8.12.6) with ESMTP id i1MLABtS000507; Sun, 22 Feb 2004 22:10:11 +0100 (CET) (envelope-from kettenis@elgar.kettenis.dyndns.org) Received: from elgar.kettenis.dyndns.org (localhost [127.0.0.1]) by elgar.kettenis.dyndns.org (8.12.6p3/8.12.6) with ESMTP id i1MLABRK000735; Sun, 22 Feb 2004 22:10:11 +0100 (CET) (envelope-from kettenis@elgar.kettenis.dyndns.org) Received: (from kettenis@localhost) by elgar.kettenis.dyndns.org (8.12.6p3/8.12.6/Submit) id i1MLAAlH000732; Sun, 22 Feb 2004 22:10:10 +0100 (CET) Date: Sun, 22 Feb 2004 21:10:00 -0000 Message-Id: <200402222110.i1MLAAlH000732@elgar.kettenis.dyndns.org> From: Mark Kettenis To: cagney@gnu.org CC: gdb-patches@sources.redhat.com In-reply-to: <4038E9A5.2090601@gnu.org> (message from Andrew Cagney on Sun, 22 Feb 2004 12:40:53 -0500) Subject: Re: [patch/rfc] The off again, on again, PC == 0 in get_prev_frame References: <40367BCD.8090403@gnu.org> <200402202323.i1KNNaL4027099@elgar.kettenis.dyndns.org> <4038E9A5.2090601@gnu.org> X-SW-Source: 2004-02/txt/msg00624.txt.bz2 Date: Sun, 22 Feb 2004 12:40:53 -0500 From: Andrew Cagney > The Linux folks should add proper CFI to the implementation of > close(2) in glibc. A DW_CFA_def_cfa rule that sets the CFA to %rbp > should do the trick. clone(2)? Yup. Sorry. Typo. The thread_start you're seeing in the backtrace is part of the clone(2) implementation in glibc. See sysdeps/unix/sysv/linux/x86_64/clone.S. How come the trad unwind code doesn't recover? There simply isn't a frame pointer. Thinking about it, the test would stop this backtrace: sigsegv_handler "null" foo (which called "null") main is this tested anywhere? No it isn't. Mark